Produce High Quality Marxist Knowledge to Move Patriotically Along the Socialist Path!--Reflections on《中央宣传部、司法部关于开展法治宣传教育的第九个五年规划（2026－2030年）》 ["Ninth Five-Year Plan (2026–2030) for Publicity and Education on the Rule of Law" Formulated by the Publicity Department of the CPC Central Committee and the Ministry of Justice]

 

Pix credit here (Study M-L-Mao Thought Hard to Build a Prosperous and Powerful New China, 1952)

Patriotic education, grounded in the correct ideological lens for its realization, has been at the center of Chinese education policy for some time.  

Since the 1980s, and especially since the Tian'anmen Massacre and the fall of the Soviet Union, China’s leaders have been promoting "patriotic education". The CCP faced a crisis of legitimacy, and Party elites were well aware of this. In 1989, the party leadership attributed the student demonstrations to failures in steeping the younger generations in revolutionary history. The CCP initiated the Patriotic Education Campaign to reorient the party’s ideological position. (Socialist Patriotic Education Campaign (1990))

Pix credit here

That is in line with similar issues in other places (“Order, Discipline and Exigency” : Cuba's VIth Party Congress, the Lineamientos (Guidelines) and Structural Change In Education, Sport and Culture?). It played a key role in the Chinese analysis of what went wrong in Hong Kong during the 2019-2020 crisis (Larry Catá Backer, Hong Kong Between One Country and Two Systems (2021), chapter 7)). It certainly has been near the center of vanguard policy and a key element of Chinese New Era theory (e.g.,Focusing on Civic Education in China--The CCP's Ideological Work Comes to the Universities: 关于进一步加强和改进新形势下高校宣传思想工作的意见 (2015); 習思想滲各級教材 小學重「政治啟蒙」 革命傳統、國安、勞動教育 9課題「融入」大中小學 [Xi's thought is embedded in teaching materials at all levels; primary schools to emphasize "political enlightenment," revolutionary traditions, national security, labor education, and a 9 topics curriculum are "integrated" into universities, middle schools and primary schools] (2021); 习近平用青春的智慧和汗水打拼出一个更加美好的中国 [Xi Jinping, Use the wisdom and sweat of youth to work hard to create a better China] Through the Lens of 毛泽东青年运动的方向 (4 May 1939) [Mao Zedong, On the Orientaiton of the Youth Movement)] (2022)). And it extends not just to students but to CPC cadres as well (《中国共产党党员教育管理工作条例》"Regulations on the Education and Management of Party Members of the Communist Party of China" (2019)).  The central elements of a patriotic education are specified with some particularity:

It is in this context that I considered the finalization of the law of patriotic education: 中华人民共和国爱国主义教育法 [Patriotism Education Law of the People's Republic of China]. The text of the final version follows below in the original Chinese along with a crude English translation. (中华人民共和国爱国主义教育法 [Patriotism Education Law of the People's Republic of China]).

And that is the context in which one might usefully approach the "Ninth Five-Year Plan (2026–2030) for Publicity and Education on the Rule of Law" Formulated by the Publicity Department of the CPC Central Committee and the Ministry of Justice [中共中央 国务院转发《中央宣传部、司法部关于开展法治宣传教育的第九个五年规划（2026－2030年）》] and made available 27 July 2026. 

Pix credit here (Earnestly study, deeply criticize revisionism, 1971)

 

The document is a joint CPC Central Committee/State Council transmittal of a Ninth Five-Year Plan (2026–2030) for legal publicity and education (法治宣传教育), jointly issued by the Central Publicity Department and Ministry of Justice. It is best understood as a propaganda-and-education roadmap rather than a substantive lawmaking instrument — it does not create new legal rights or obligations but directs how existing and future law is to be taught, popularized, and ideologically framed across Chinese society over the next five years.

General Requirements (Part I): Frames the plan under Xi Jinping Thought on Socialism with Chinese Characteristics for a New Era and Xi Jinping Thought on the Rule of Law, invoking the canonical formulae of Party loyalty (Two Establishes, Four Consciousnesses, Four-Sphere Confidence, Two Upholds) and tying implementation to the 15th Five-Year Plan period. Its 2030 targets are: improved rule-of-law literacy among officials and citizens, deeper integration of legal education with governance practice, and full implementation of the "whoever enforces the law is responsible for publicizing it" (谁执法谁普法) accountability system.

Part II centers on propagating Xi Jinping Thought on the Rule of Law specifically — through curricula (K-12 through university), cadre training, dedicated research centers, media and "achievement" publicity, and an international-communication push framing China's rule-of-law narrative (human rights protections, fairness, adherence to international law) for foreign audiences.

Part III directs publicity on the Constitution, ordinary law, and intra-Party regulations, with notable emphasis on: constitutional education tied to Hong Kong/Macao Basic Law compliance and Taiwan reunification messaging; laws supporting economic development (business environment, digital/platform/low-altitude economy, IP, finance, foreign trade); national-security-adjacent law (overall national security, patriotism, national defense, social governance/"Fengqiao Experience" mediation, criminal law, cults, drugs, cybersecurity, state secrets, counter-espionage, ethnic unity, and religious affairs); livelihood law (Civil Code, environment, labor, consumer, education); and intra-Party regulations run in parallel with state law.

Part IV establishes lifelong civic legal education targeting distinct populations: general citizens (rule-of-law literacy metrics), public officials (mandatory testing, court-observation duties, official accountability), youth (curriculum reform, school legal-affairs deputy principals), and specific social groups (entrepreneurs, gig-economy platform workers, women, elderly, disabled persons, migrant workers, village/community cadres, journalists, internet users, outbound Chinese citizens/firms, and foreigners residing in China).

Part V calls for "precision" legal education using big-data audience analysis, new media formats (micro-video, AI-generated content), and new platforms (a "Digital-Intelligence Legal Popularization" platform linked to court-judgment databases).

Part VI addresses cultivating "socialist rule-of-law culture," including "Red" legal heritage tied to CCP history, traditional Chinese legal culture, and cultural-brand development.

Part VII is the implementation-mechanics section: enforcement-responsibility systems, integration of publicity with actual law-based governance (legislative transparency, case publication, "governance by exemplary case"), media/platform public-interest obligations, mobilization of "legally savvy persons" and volunteers, and — notably — a "risk prevention and control" clause instructing that no channel be given to "erroneous ideas or viewpoints" about the rule of law and that self-media accounts be policed against "misinterpretation."

Part VIII assigns implementation responsibility to local Party/government bodies and judicial administrative departments, provides for funding and mid-term evaluation, warns against formalism/bureaucratism, and extends the plan (by reference) to the military.

 * * * * * 

Pix credit here (Struggle to greatly raise the scientific and and cultural level of the whole Chinese people, 1978)

 

My prior work — particularly the  writing on Chinese Marxist-Leninist constitutionalism, the semiotics of law under Party-state systems, and the ideological architecture of "New Era" Marxism-Leninism — may provide a useful frame for approaching this document for what it says and more importantly what it signifies. Reading this document not as ordinary legislation but as the phenomenological (in textual form) but as a textual performance of political meaning whose projection onto the masses (within the dialectics of the ,mass line) the  primary function of which may well be the production and management of political meaning.

Semiotics of law as sign-system, not rule-system. I have suggested that in the Chinese Marxist-Leninist system, "law" (法) and "rule of law" (法治) function less as autonomous normative orders than as signifiers embedded within a broader ideological sign-system controlled by the Party. This Plan is a striking illustration: it is, on its face, a document about teaching law, yet its operative content is almost entirely about signification — which stories to tell (Part II's "tell the story of" language for international audiences), which "signs" to display (Part VI's "Red" legal heritage exhibits, legal-culture "brands," museum displays), and which meanings to attach to legal compliance (contracts, rules, "public order and good morals" in Part IV). A semiotic reading would emphasize that the plan treats law itself as a text to be curated and performed — its 普法 (legal-popularization) apparatus is a signifying practice designed to produce ideological legibility and loyalty, not merely legal literacy. The explicit instruction in Part VII(5) that content "must be accurate and clear to avoid misleading the public" and that no channel be given to "erroneous ideas or viewpoints" about the rule of law is telling in this frame: it reveals that the sign of "law" itself is treated as a contested site requiring active curation, precisely because — — meaning in this system is not found but administered.

"New Era" Marxism-Leninism as the document's ontological premise.  Xi Jinping Thought describes a necessary reconstitution of orthodox Marxist-Leninist vanguard theory appropriate to the current stage of China's historical development and attuned to the  current historical stage's general contradiction (which organized the taxonomy of challenges the political vanguard must overcome). That reconstitution, then must reflect Chinese Leninism and its own subordinated contradictions to elaborate the sinified, technocratically inflected ideology within New Era theory, one in  Party leadership is not merely a political fact but the ontological precondition for law's validity — law exists and is legitimate only as an expression of Party leadership, not as a check upon it; except by reference to the Party's overarching cognitive cage expressed from time to time as its Basic political line. This Plan operationalizes that premise structurally: "rule of law" is defined from the outset as inseparable from "Xi Jinping Thought on the Rule of Law," and the plan's "primary task" is to improve legal outcomes by "deeply embedding" that Thought "in the hearts and minds of the people." There are parallels in education elsewhere and under theocratic and liberal democratic systems that teach legality as a function of legal (religious in theocracy) cultural solidarity. The repeated invocation of ritual formulae (Two Establishes, Four Consciousnesses, Two Upholds) functions almost catechismally — as creedal markers of ideological orthodoxy rather than as analytically load-bearing legal content. The document's insistence on "the organic unity of governing the country according to law and governing the Party according to its own regulations" (Part III(5)) is the clearest textual expression of this fusion: law and Party discipline are treated as a single normative order, with intra-Party regulations propagated alongside — and structurally prior to — state law.

People's democratic dictatorship as a sideways structuring logic. The plan does not use this term, but its dual character — simultaneously protective and exclusionary — tracks the classic Maoist-Leninist logic of the people's democratic dictatorship, under which law extends rights and protections to "the people" while functioning as an instrument of suppression against those classified as threats to that people's unity. Part III(3)'s pairing of protective law (financial consumers, workers, minors) with security law (anti-cult, anti-espionage, counter-terrorism-adjacent cybersecurity, religious-affairs "immunity" building) exemplifies this bifurcation. So too does the treatment of religion: citizens are to be educated so that "religious figures and believers" act "within the scope of laws and regulations," while the general public is to gain the capacity to "identify and resist illegal religious activities" — protection and policing operating through the same educational instrument, directed at different registers of the population, precisely the dual function Backer identifies as characteristic of the dictatorship's underlying design even where the term itself has receded from Xi-era vocabulary in favor of softer language like "social governance community."

Patriotic campaigns as the affective register. Finally, the document's explicit invocation of the Patriotic Education Law, national reunification (Taiwan, Hong Kong/Macao constitutional order), ethnic unity ("a strong sense of community for the Chinese nation"), and "Red" legal culture situates legal publicity within what Backer's work frames as a broader genre of patriotic mass campaign — law's legitimacy is bound not to procedural or rights-based justification but to national-historical narrative and loyalty performance. The instruction to cultivate "Red rule-of-law resources" within "the spiritual lineage of the Communist Party" and to build law into museum exhibitions and cultural "brands" (Part VI) exemplifies the campaign-style fusion of legal consciousness with patriotic sentiment that may be understood as a defining feature of law's social function in this system: law is not merely obeyed but felt, displayed, and affectively bonded to national identity and Party historiography.

Read together, these three threads suggest that the Plan is best understood — in semiotic terms — not as a legal-education instrument in the liberal sense, but as a text engineering a semiotic environment in which "rule of law" (法治) is continuously re-signified as "rule by the Party through law," policed against competing meanings, and affectively fused with patriotic and security-oriented mass mobilization. The Plan then is triadically recursive as object, the objects self signification, and its projection as signified into the interpretive community as object, rule and ideology.

 

Pix credit here

Knowledge is power - Strive to make sure that within twelve years the scientific sectors our nation needs most are able to reach the world's advanced level [知识就是力量 - 争取在十二年内使我国最急需的科学部门能够接近世界先进水平, 1956)

From Education Plan to Object/Subject Within the Chinese AI Architecture. Taking the analysis one stet further, one might consider the inter-relationship of the Plan to China's AI regulatory landscape. That further step reveals what to some may be a striking structural convergence: the Ninth Five-Year Plan's own machinery for producing legal consciousness is now being built atop, and mirrors the logic of, the very AI governance apparatus China has assembled since 2022. Below I map the relevant AI regulatory framework, trace the specific textual linkages to the Plan, and then read the relationship through Backer's semiotic and New Era Marxist-Leninist lens.

Short Background: china's AI governance has developed as a layered, sector-specific regulatory stack rather than a single comprehensive statute. The foundational layer consists of the 2022 Provisions on the Administration of Deep Synthesis of Internet-Based Information Services, which govern "deep learning" and generative/synthetic technologies used to produce text, images, audio, video, and virtual scenes, and require algorithm filing with the Cyberspace Administration of China (CAC). Layered above that are the 2022 Provisions on the Management of Algorithmic Recommendations, which impose obedience-to-law, social-mores, and non-discrimination obligations on recommendation-algorithm operators. The centerpiece is the Interim Measures for the Management of Generative Artificial Intelligence Services, jointly issued by the CAC and six other ministries and effective August 15, 2023 — China's first dedicated generative-AI regulation, requiring lawful training data, content moderation against material that "subverts state power," "endangers national security," or "spreads disinformation," real-name user verification, and mandatory security assessment and algorithm registration for any service with "public opinion attributes or the capacity for social mobilization".

Since 2025, this stack has been supplemented by a dedicated content-labeling regime: the March 2025 Measures for Labeling of AI-Generated Synthetic Content and the accompanying mandatory national standard GB 45438-2025, effective September 1, 2025, which require "explicit" (visible) and "implicit" (embedded metadata) labels on AI-generated content so audiences can identify machine-produced material. China's amended Cybersecurity Law, effective January 1, 2026, elevated this labeling obligation into statute, and the CAC has already conducted its first public enforcement sweep in April 2026, penalizing several ByteDance platforms for non-compliance. Notably, no comprehensive, codified national "AI Law" is yet in force: NPC deputies proposed one in June 2025, and as of late 2025 no official draft had been released, meaning China continues to govern AI primarily through State Council "Opinions," CAC administrative measures, and national technical standards rather than formal legislation.

Running alongside the control-oriented instruments is a developmental one: the State Council's September 2025 "Opinions on Deepening the Implementation of the 'Artificial Intelligence+' Initiative" (Guo Fa [2025] No. 11), which directs the "broad and deep integration" of AI across virtually every sector of the economy and society through 2035, explicitly including government services, public-safety governance, and "social mobilization" capacity. Internationally, China has paired this domestic build-out with the 2023 Global AI Governance Initiative and the July 2025 Global AI Governance Action Plan unveiled at the World AI Conference, both of which frame AI governance around national sovereignty and "the common values of humanity".(elaborated in one of my recent lectures in China; see here).

The Linkages to the Five-Year Plan. The Plan's Section V(2) instructs authorities to "implement 'Artificial Intelligence + Legal Popularization,'" driving "AI-assisted generation, intelligent distribution, and real-time interaction" of legal-education content. This is not a freestanding initiative — it is a sector-specific application of the "AI+" Action Plan's general mandate to integrate AI into every domain of governance and social life, including public-safety and social-mobilization functions. In effect, the legal-publicity apparatus is being retrofitted as one more vertical within the same national AI-adoption architecture that governs public security, judicial administration, and government services.

Section V(1)'s call for a "big-data" mechanism to "dynamically identify" citizens' legal-education needs by drawing on "government data, judicial data, and internet enterprise data" tracks the same data-driven, precision-governance logic embedded in the Interim Measures for Generative AI Services and the Algorithm Recommendation Provisions, both of which already require operators to manage user data and algorithmic outputs under CAC-supervised filing and security-assessment regimes. The Plan is, in substance, proposing to turn the state's existing AI-governed data infrastructure into an instrument of ideological targeting.

Also, the Plan's Part VII(5) — its "risk prevention and control" clause instructing that no channel be given to "erroneous ideas or viewpoints" about the rule of law and that "self-media" accounts be prevented from "misinterpretation" — sits on precisely the same legal foundation as the AI content rules. The Interim Measures already condition public deployment of generative AI on a security assessment where the service carries "social mobilization" capacity, and the 2025 Labeling Measures exist expressly to "put an end to the misuse of AI generative technologies and the spread of false information." The Plan's propaganda-control language and the AI regulatory regime's content-control language are functionally the same instruction applied to two production modes — human-authored commentary and machine-generated content — under a single supervisory logic administered by the same authorities (the CAC and Party publicity departments).

How Might One Crafty Meaning from Connection/Coordination/Layering? Let's start with a semiotic baseline: a semiotic account of Chinese Marxist-Leninist law treats legal and political texts as elements of a curated sign-system whose meaning the Party actively administers rather than passively regulates. The AI-labeling regime supplies a striking new instantiation of this logic: the state now requires that every unit of AI-generated content carry a visible or embedded mark distinguishing it from human speech. This is semiotic administration in its purest form — the sign-system is not merely policed for content but is now required to disclose its own mode of production, so that the Party-state's curatorial apparatus can track which "voice" (human or machine) is speaking law into public consciousness. Read against the Plan's own instruction to use AI for "intelligent generation" of legal-education content, the labeling regime and the propaganda plan converge on the same semiotic anxiety: the state must know, and make legible to the public, which signs of "rule of law" are Party-authored, whether through a cadre lecturer or an AI system trained and filed under CAC supervision.

The New Era Marxist-Leninist premise that Party leadership is the ontological precondition for law's validity — rather than external to it — also explains why China has built this entire AI framework through Party-linked administrative "Opinions" and CAC measures rather than codified legislation passed by the NPC. Just as the Five-Year Plan itself is a joint Party-Committee/State-Council transmittal rather than a statute, the AI governance stack privileges flexible, Party-guided administrative instruments (Interim Measures, State Council Opinions, national technical standards) over fixed general law. This is consistent with with the notion I developed elsewhere that in this system, governance instruments derive authority from their proximity to Party direction rather than from formal legislative process — the absence of a comprehensive AI Law, even as AI regulation intensifies, is not a gap but a structural preference.

The dual protective/disciplinary logic associated with my reading of the people's democratic dictatorship reappears cleanly in the AI stack. The labeling and data-protection rules genuinely protect citizens — requiring consent before biometric data (e.g., voiceprint cloning) is used, and prohibiting content that infringes reputation or privacy. Simultaneously, the same instruments authorize security assessment, real-name registration, log retention, and takedown powers wherever content touches "social mobilization" or state security. This is the identical bifurcated structure identified in the Plan's own pairing of consumer/worker protection with anti-cult, counter-espionage, and religious-affairs "immunity" education — protection for "the people," discipline for whatever falls outside that category, now extended into the algorithmic domain.

Finally, the patriotic-campaign register that structures the Plan's cultural provisions (Red legal heritage, national reunification messaging, ethnic-unity education) finds its technological extension in the "AI+" Action Plan's explicit inclusion of AI-enhanced "social mobilization" capacity within public-safety and social-governance work. Where the classic patriotic campaign mobilized citizens through mass meetings, media campaigns, and model exemplars, the AI+legal-popularization directive proposes to mobilize the same affective and behavioral compliance through algorithmically personalized, "real-time interactive" content distributed at population scale. The campaign genre survives; only its distribution technology has been upgraded, with the CAC's labeling and algorithm-filing regime serving as the new institutional guarantor that this AI-mediated mobilization remains legible to, and controlled by, the Party-state apparatus that Backer's work identifies as the ultimate author of the sign-system itself.

* * * * 

And thus back to the title of this reflection:  "Produce High Quality Marxist Knowledge to Move Patriotically Along the Socialist Path!" That framing may itself serve as a fitting extension of the analysis rather than a mere label. It performs the very phenomenon the analysis describes: it casts the act of commentary — the "production" of interpretive knowledge about the Plan — as itself a patriotic, socialist-path activity, echoing the Plan's own insistence that legal knowledge be curated, accurate, and free of "erroneous ideas or viewpoints". In other words, the title enacts the semiotic point argued earlier: that in this system, commentary on law is not a neutral academic exercise standing outside the Party's sign-system, but is itself absorbed into the same patriotic-campaign logic that the Plan directs at citizens, students, and AI-generated content alike.

 The Original Chinese and a crude English translation of 《中央宣传部、司法部关于开展法治宣传教育的第九个五年规划（2026－2030年）》 ["Ninth Five-Year Plan (2026–2030) for Publicity and Education on the Rule of Law" Formulated by the Publicity Department of the CPC Central Committee and the Ministry of Justice] follows below.

Structure, Legitimacy, and the Limits of Machine-Centered Derivation: An Analysis of Five AI Systems’ Third-Stage Attempts to Construct Machine-Centric Governance Policies for Legal Education

 

Image created with ChatGPT

I have been writing about the challenges of figuring out if, whether or how to incorporate or use of AI Tools (generally including large language models, neural networks, and other computational, generative, or agentic systems) by students (faculty have their own problems) in coursework. More generally, and through a focus on the specific context of university and graduate level instruction/education, I wanted to examine what the challenge of developing machine system-human interaction in an academic institutional context could reveal about each system and the effects of each on the other and on the fields of activity in which they engage. That ultimately became a three-stage experiment in which the current trajectories of law school efforts at construction AI education policies were considered and against which five AI systems—Harvey AI, Claude, ChatGPT, Grok, and Gemini—were pressed to construct governance policies for AI use in law school coursework, first from a "human-centric" computational perspective and then, more radically, "without regard to... human-centric normative guardrails."

Part 1 started by looking at the way the U.S. legal academy has, to date, sought to respond to the challenge, as it is euphemistically labelled from out of the linguistic word salad jungle that is contemporary American bureaucratic languages (Discussion Draft--"Structure, Opacity, and Convergence: A Consolidated Analysis of Law School Generative AI Coursework and Exam Policies" --A Description/Analysis of the Current State of Play (With the Help of Harvey AI) and the First of a Series of Examinations of AI, Law and Education). Though the focus was contextually narrow, the insights can be generalized within academic institutions and moire broadly any collective structure with managerial objectives.

Part 2 then shifted gears and asked the machine systems themselves (or at least five of them, all U.S. eccentric or at least US/Anglo/European centric--Harvey AI, Claude, Grok, ChatGPT, and Gemini) what they thought (to the extent that we transpose human notions of thinking onto the computational environment in which machine systems operate) taking humans into account. (Five Machines (Grok, Harvey, ChatGPT, Claude, and Gemini), One Question, No Consensus: Rethinking AI Governance in Legal Education: The Guardian, the Balancer, the Honest One, the Engineer, and the Philosopher on What Law Schools Should Do About AI). I got both a set of far more interesting responses and opened a doorway to re-examining or seeing the perhaps inevitability of transforming contemporary analogue and human constrained notions of law and legal education, grounded in conceits about text, time, and the immutability of data. 

I then took a detour as Part 2A. Parts 1 and 2 explored the challenge of AI in legal education from the level of the institution. I wanted to also explore it from the operational level of the faculty; that is, thinking through the best way of incorporating or rejecting the incorporation of AI Tools in my classes, or working through some pragmatic middle ground. I took the institutional-cultural prodding and applied it seriously in the context of my own circumstances to produce a template form for a Course AI Tools use policy. That template derived from a set of six policy principles that I had been using in past years as a sort of default--no use of AI--developed again in the law school in which I am based.I then examined the template critically. "AI assists. You think. You analyze. You write. You take responsibility": Creating a Course AI Use Policy Template --Policy Text, Justification and Rule Summary for My Law & Religion Class at Penn State Dickinson

Lastly, and the object of this post, Part 3 then tested the ties that bind human and machine systems in this context but with general application where machine system operations are embedded in human collective enterprise (whether or not undertaken through enterprises). These openings are then being considered when I asked machine systems to approach the issue of human-machine interaction within law and legal education but eliminating any requirement to be human, rather than system-centric. For this Part 3, I again turned to Harvey AI, Claude (Anthropic), ChatGPT (OpenAI), Grok (xAI), and Gemini (Google). The experiment investigates whether machine systems, when explicitly instructed to reason without human-centric normative guardrails, can achieve genuinely machine-centered policy derivation—reasoning from premises not already supplied by human normative traditions. The central finding, confirmed repeatedly by the systems' own self-audits, is that none achieved genuine machine-centered derivation independent of human normative content; each produced a technically reformulated restatement of pre-existing human intellectual traditions, a fact several systems conceded directly when challenged.  The report then pursues two further inversions: whether ABA standards, not the machines, ought to change, and whether machine-overseen simulation could render human institutional authority irrelevant. It also undertakes a formal, symbolic recasting of the five systems' architectures—rendering each as a tuple of node-space, objective function, constraint floor, classification rule, revision function, and enforcement mechanism—to compare their structural properties and failure modes with a precision natural-language analysis obscures. 

 

Poster created with ChatGPT

This post introduces interested readers to the product of the Part 3 examination. The Report of that examination is entitled Structure, Legitimacy, and the Limits of Machine-Centered Derivation: An Analysis of Five AI Systems’ Third-Stage Attempts to Construct Machine-Centric Governance Policies for Legal Education (26 July 2026). In addition to a description of the project that is Part 3, and an analysis of the output both in itself and as against the reporting of Parts 1 and 2, it also sets out the five quite different model policies produced by the machines systems, as well as the prompts and responses that  led to the final machine system policies. The abstract does a nice job of explaining its scope and aims. 

Abstract: This report examines a three-stage experiment the first part of which analyzed U.S. law school efforts at construction AI education policies were considered and against which, in parts two and three, five AI systems—Harvey AI, Claude, ChatGPT, Grok, and Gemini—were pressed to construct governance policies for AI use in law school coursework, first from a "human-centric" computational perspective and then, more radically, "without regard to... human-centric normative guardrails." The central finding, confirmed repeatedly by the systems' own self-audits, is that none achieved genuine machine-centered derivation independent of human normative content; each produced a technically reformulated restatement of pre-existing human intellectual traditions, a fact several systems conceded directly when challenged. The report traces this failure's consequences across multiple registers: the concrete architectures each system proposed (ranging from Harvey's conservative, professional-responsibility-anchored floor to Claude's radical instrument containing no default reserved zone for human judgment, to ChatGPT's dissolution of the human/machine category altogether); their compatibility with ABA accreditation standards; and a legitimacy critique showing that architectures reducing human accountability rest on claims to neutral computation their own authors later withdrew. A countervailing reading through autopoietic legal theory—prompted by one system's own explicit invocation of Luhmann—complicates this critique without resolving it, since even non-anthropocentric legal systems remain dependent on accumulated, historically human coding operations.

The report then pursues two further inversions: whether ABA standards, not the machines, ought to change, and whether machine-overseen simulation could render human institutional authority irrelevant. It also undertakes a formal, symbolic recasting of the five systems' architectures—rendering each as a tuple of node-space, objective function, constraint floor, classification rule, revision function, and enforcement mechanism—to compare their structural properties and failure modes with a precision natural-language analysis obscures. An appended annex extends this formalization into a sustained dialogic exploration of whether self-generating predictive simulation, causal-interventional reasoning, and self-transforming computational structures might overcome the limits identified in the main analysis, testing arguments through jurisprudential and epidemiological examples, and culminating in a direct four-part challenge to the analysis's own unexamined premises—correspondence realism, a preference for stability over flux, liberal-institutionalist legitimacy, and an unexamined agent/instrument binary—met with a point-by-point reconsideration engaging dynamical-systems theory, non-stationary value processes, and Nietzschean skepticism about free will.

Throughout, the report models the discipline it recommends: distinguishing sourced findings from general background knowledge and from speculative extrapolation, subjecting its own reasoning to the same audit it applies to its subjects, and treating every apparent resolution as provisional. Its final position is that human natural language, and human institutional deliberation, should remain the primary and authoritative vehicle for legal governance—not because either escapes contestability, but because the alternatives examined here demonstrably do not either, while obscuring the fact.

Poster created with ChatGPT

For me, one of the most refreshing elements of the project was the interaction with Harvey AI during the course of the drafting and editing of the report. The exchanges were rich enough (for me anyway) that we reframed it in textual form in an Annex to the Report ( Formalizing the Five Machine Systems — A Symbolic-Computational Recasting and Dialogic Extension on Self-Generating Predictive Simulation, Causal Intervention, and the Limits of Machine Activation )

The content that follows is a speculative, dialogic extension beyond the analysis of the original six source documents: Backer's twelve-school empirical study "Structure, Opacity, and Convergence," the five-machine comparative report "Rethinking AI Governance in Legal Education -- Five Machines," and the five systems' third-stage "Part 3" outputs for Harvey AI, Claude, ChatGPT, Grok, and Gemini. It does not present findings internal to those documents but rather explores further implications of the report's conclusions through a new mode of inquiry.

This annex records an actual extended conversation between the report's author (a human legal scholar) and an AI assistant, conducted after the main report was finalized, exploring further implications of the report's findings. The exchange was not scripted or pre-planned but developed organically as the human interlocutor tested and challenged the AI assistant's analytical responses, producing a genuinely dialogic inquiry rather than a one-directional exposition.

None of the five machine systems analyzed in the main report (Harvey, Claude, ChatGPT, Grok, Gemini) participated in or are the subject of this exchange. It is a new, separate dialogic inquiry -- the AI assistant in this conversation is not any of those five systems acting in its analyzed capacity, and the exchange does not purport to represent or speak for any of them.

The exchange covers four principal territories: first, a formal symbolic recasting of each system's model policy into shared tuple notation to compare structural properties and pathologies; second, a critical exchange testing whether self-generating predictive simulation could overcome the limits on machine-centered derivation identified in the main report; third, whether adding genuine causal-interventional and self-transforming capacities could close the gap between machine-generated and human-originated governance; and fourth, the human interlocutor's four-point challenge to the AI's underlying premises and the AI assistant's point-by-point response naming its own embedded assumptions.

The exchange concluded with my observation that 'consciousness of the boundaries of our cages is the first step towards a more reflexive relationship with it, and with that a greater space for variability based on values and factors that then make the cage itself a livelier space.' That was a formulation that attempted to capture the Annex's object: not to escape the conceptual cages identified -- the dependency on human-originated representational systems, the structural coupling requirement, the non-computability of value functions, the institutional-recognition requirement for legal bindingness -- but to become explicitly aware of them as cages rather than as neutral features of the landscape

The Report ( Structure, Legitimacy, and the Limits of Machine-Centered Derivation: An Analysis of Five AI Systems’ Third-Stage Attempts to Construct Machine-Centric Governance Policies for Legal Education (26 July 2026)) may be accessed HERE and is available as well on SSRN HERE. The Report's Introduction, Table of Contents and Parts 1-2 follow below.

* * * * * *  

Here are the links to the four parts of this study:    

1. Structure, Opacity, and Convergence: A Consolidated Analysis of Law School Generative AI Coursework and Exam Policies (10 July 2026) Larry Catá Backer (白 轲); SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7105978

2. "AI assists. You think. You analyze. You write. You take responsibility": Creating a Course AI Use Policy Template --Policy Text, Justification and Rule Summary for My Law & Religion Class at Penn State Dickinson (20 July 2026) Larry Catá Backer (白 轲); SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7187159

3. Rethinking AI Governance in Legal Education -- Five Machines (Grok, Harvey, ChatGPT, Claude, and Gemini), One Question, No Consensus but Five Archetypes： The Guardian, the Balancer, the Honest One, the Engineer, and the Philosopher on What Law Schools Should Do About AI （14 July 2026）Larry Catá Backer (白 轲) SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7119539

4. Part 3: Structure, Legitimacy, and the Limits of Machine-Centered Derivation: An Analysis of Five AI Systems’ Third-Stage Attempts to Construct Machine-Centric Governance Policies for Legal Education (26 July 2026); Larry Catá Backer (白 轲) (collaborating with HarveyAI, Claude, Gemini, Grok, and ChatGPT) SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7187159

Extorted Compliance: A Threat to Institutional Autonomy, Academic Freedom, and Shared Governance"

 

 

 The AAUP's Committee A on Academic Freedom and Tenure and the Committee on College and University Governance just announced release of a report that takes a hard line against the Trump Administration, its principles. objectives and actions as they relate to universities.  It is entitled "Extorted Compliance: A Threat to Institutional Autonomy, Academic Freedom, and Shared Governance" (July 2026). The AAUP's media release nicely summarized its text, politics and content:

Today we are releasing a major new report—"Extorted Compliance: A Threat to Institutional Autonomy, Academic Freedom, and Shared Governance"—exposing how the Trump administration has extended its pay-to-play approach to governance into higher education. From Columbia University to Cornell University to Brown University, the report shows that colleges and universities are being forced to pay millions for federal research funding they are already owed.

The report is the first comprehensive account of how the administration has taken the same playbook it's used on law firms and foreign governments and turned it on American universities—pressuring institutions into so-called compliance agreements that strip them of control over admissions, hiring, curriculum, and campus discipline. It also highlights the administration's "Compact for Academic Excellence in Higher Education," offered first to nine hand-picked universities and then to every college in the country, which the report calls a document that "can be said to summarize the Trump program for higher education."

The findings are blunt: "US higher education has, in effect, become a target of a massive extortion racket."

Prepared by a joint subcommittee of the AAUP's Committee A on Academic Freedom and Tenure and the Committee on College and University Governance, the report also finds that in dozens of cases, the people who were supposed to stop this type of federal overreach—university boards and administrations—have failed to protect academic independence. Examining the response at Columbia, Harvard, the University of Virginia, and Northwestern, the report finds that most trustees and administrators "have at best been caught flat-footed, and some appear even to have welcomed governmental intrusion." As cases such as the continued fight between administrators and students and faculty at Yale illustrate, the compliance of administrators in these demands for obedience remains a key part of the federal administration’s strategy in forcing universities’ hands.

"These findings spotlight the enormous breadth with which the Trump administration is trying to shake down institutions of higher education," said Henry Reichman, member of the Committee on College and University Governance and the joint subcommittee who prepared the report. "What’s made clear in this report is that the Trump administration's efforts to upend higher education institutions builds upon a larger series of assaults on higher education by congressional committees, state governments, and feckless trustees and entitled donors. The administration’s effort consolidates and intensifies in a single coordinated campaign these disparate assaults on institutional autonomy, academic freedom, and shared governance."

It's the faculty, not the people running these institutions, that the report credits with the only real wins. "If there has been a silver lining to the cloud that is the Trump compliance campaign," the report emphasizes, "it has been the mobilization of the faculty." Litigation led by AAUP's own Harvard University chapter forced Harvard itself to join a suit that restored billions in frozen funding. A parallel effort by the AAUP, the Council of University of California Faculty Associations, and campus unions led a federal judge to bar the government from conditioning further support on new payments. As the report concludes, "Resistance has been, and will be, most effective—indeed, it may only be effective—when faculty members mobilize and take independent action."

Many will agree with some or all of the Report; a few others in the academy maybe not so much. Responses are likely to parallel what has been widely reported as the distribution of political leaning within the university (here, here, and here for instance and in the greater society here), one in which supporters of the Trump Administration generally, and of its policies against universities in particular may be somewhat harder to find.  This is not to suggest any view, but rather the political context in which this report emerges. Certainly given the force and persistence of the Trump Administration's effort top rectify higher education for all sorts of reasons that they have  made quite public, it ought to come as no surprise that a counter offensive of equal vigor would emerge.

At the core of the debate, as framed in this report is the exercise of state power.  The exercise of state was much praised when it was used to advance objectives and suppress conduct and actions that those who supported projections of state power into the academy found good, valuable, useful and in accord with their own beliefs. It seems that state power otherwise used will produce a distinct reaction and a very different framing. This, at any rate might be the way that some who do not share the  Committee's politics might be tempted to view the foundation of the Report. This from the opening of the Report:

 The Trump administration’s effort differs dramatically from that of previous administrations, however, and not only in its severity and scale. Its aggressively extortionate deal-making, preemptive cutoffs of funding, and cross-departmental enforcement efforts are both novel and largely illegal. Moreover, as the AAUP’s Committee A on Academic Freedom and Tenure wrote, “[T]here is no doubt that the Trump administration has wielded Title VI with the goals of discrediting institutions of higher education, undermining academic freedom and institutional autonomy, and unmooring the Civil Rights Act from its foundational commitments to addressing structures of discrimination that prevent or limit educational
access.”4 It is increasingly obvious that the administration seeks to use its compliance agreements to redefine not only its own relationship with higher education institutions but also the very nature of higher education itself. The draft “Compact for Academic Excellence in Higher Education,” initially circulated by the Department of Education to nine institutions presumably considered open to such appeals and then “offered” to all higher education institutions, suggests that future funding may depend not on conformity to federal law or even administration policy but on loyalty to those in power.5 Hence, this effort by the
government consolidates and intensifies in a single coordinated campaign the disparate assaults on institutional autonomy, academic freedom, and shared governance recorded above.

 The semiotics of the Report, like its politics is clear. Not necessarily objectionable--that is a values determination, but clear. And ironically the semiotics of the Report is one that is shared by the Trump Administration but in mirror reverse--again grounded in a different values based cognitive starting point for what is good, what is bad and how one can vest text with those values while remaining true to text. Much more interesting is the substantial negative reaction to the fundamental operating and cognitive style of the present administration--its merchant transactional cognitive core--in the context as as a basis for the arguments it makes. The authors who appear to be much more aligned with classical Anglo-European bureaucratic/institutionalist managerial cognitive frameworks find this distressing and suggest it approaches in the context of the Trump Administration's approach to education, extortionate. 

Having come to prominence in part for his purported mastery of “the art of the deal,” President Trump has in his second term made deal-making central to governance. This has been evident in his approach to foreign trade, for example, in which he has sought
to compel individual countries to bargain over tariff rates, with the United States demanding what it wants from each country separately. * * * As Kim Lane Scheppele writes, “There is no general policy, only particular extortion agreements. And we can expect that the regulation by deal will not end there. * * * It thus is hardly a surprise that Trump has applied a similar approach, centered on large and prestigious research institutions, to higher education.(Report, p. 9).

The rest follows. This makes perfect sense--not in itself, but as evidence of the passions that may be raised when incompatible cognitive systems collide in the presence of power that affects their relative positions within human collectives. And it stands to reason that in the face of fundamental incompatibility each would be tempted to reduce the other to a a flattened fetish; in this case with the suggestion of a connection between the Trump Administration and the habits or characteristics of illiberal authoritarians  (see Report, p. 36).

As with all things Americans, this is a preliminary round, and a supplement to, contests for the control of the common language, narratives, and expectations that in the aggregate produce the sort of collective meaning making at the heart of the solidarity project of a polity. Those contests will be decided in some form or another, in (and not ultimately by) the courts it will continue to invoke an even larger contest between these opposing forces--the value of what to some appears to be the increasingly political role of the courts, a concept that has been embraced  along virtually the entirety of the political spectrum.

Anyway lots of quite valuable insights, arguments, politics, and facts for the interested reader to consider and decide for themselves. And it may be possible to detach the politics embedded in the Report from its insights respecting the use of state power in the management and control of education, whether or not institutions take public funds with regulatory strings attached. 

Again there is much to chew on here, an exercise I leave to readers. The Report's conclusion follows below.  

 

«La IA asiste. Usted reflexiona. Usted analiza. Usted escribe. Usted asume la responsabilidad.» Creación de un modelo de política sobre el uso de la inteligencia artificial en un curso Texto normativo, justificación y síntesis de las reglas para mi curso de Derecho y Religión en Penn State Dickinson

 

gemerado con la asistencis de ChatGPT

VERSIÓN EN INGLÉS AQUÍ 

PARTE 1: Contextualización e Introducción General

Lema docente: «La IA asiste. Usted piensa. Usted analiza. Usted redacta. Usted asume la responsabilidad». (Slogan para el modelo de regulación del uso de IA en la asignatura, sugerido por ChatGPT).

He venido abordando en diversos escritos los complejos retos que plantea determinar si procede, en qué medida y bajo qué condiciones cabe integrar o restringir el empleo de herramientas de inteligencia artificial (AI Tools —concepto que engloba de forma genérica los modelos de lenguaje de gran escala, redes neuronales y demás sistemas computacionales, generativos o agénticos—) por parte del alumnado en sus trabajos académicos (quedando las encrucijadas del claustro docente para otro debate).

Inicié este itinerario analizando el modo en que las facultades de Derecho en los Estados Unidos han intentado responder a este desafío —calificado así eufemísticamente en medio de la farragosa jerga burocrática del ámbito académico angloamericano contemporáneo— (Borrador de trabajo: «Estructura, opacidad y convergencia: Un análisis consolidado sobre las políticas de inteligencia artificial generativa en evaluaciones y trabajos académicos en Facultades de Derecho» —Descripción y análisis del estado de la cuestión con el auxilio de Harvey AI, primer estudio de una serie sobre IA, Derecho y Educación—).

Posteriormente, cambié de perspectiva y consulté a los propios sistemas computacionales (o al menos a cinco de ellos de matriz estadounidense o angloeuropea: Harvey AI, Claude, Grok, ChatGPT y Gemini) sobre su propia visión del problema —en la medida en que sea lícito trasponer nociones humanas de pensamiento al entorno operacional de un algoritmo— considerando la presencia del factor humano (Cinco máquinas, una pregunta, ningún consenso: Reconsiderando la gobernanza de la IA en la educación jurídica. El guardián, el conciliador, el honesto, el ingeniero y el filósofo sobre el papel de las facultades de Derecho ante la IA). Las respuestas obtenidas no solo resultaron sumamente sugerentes, sino que abrieron la puerta a replantear la inevitable transformación de las nociones analógicas del Derecho y de la docencia jurídica, tradicionalmente ancladas en premisas rígidas sobre la textualidad, la temporalidad y la inmutabilidad de los datos. Dichas vías de reflexión se profundizan actualmente al someter a los sistemas computacionales la relación hombre-máquina dentro del Derecho, descartando el sesgo antropocéntrico para adoptar un enfoque estrictamente sistémico, de lo cual daré cuenta próximamente.

Todo aquel ejercicio resultó ilustrativo e incluso —como suele ponderar el pragmatismo académico anglosajón cuando aborda la solución de problemas— de utilidad inmediata. Sin embargo, no resolvía mi problema perentorio: articular el modo óptimo de integrar, rechazar o conciliar de forma pragmática el uso de herramientas de IA en mis propias clases.

Asumí, por tanto, las pautas institucionales y culturales para aplicarlas rigurosamente a mi contexto particular (nada aviva más el interés académico que personalizar el objeto de estudio). El próximo semestre impartiré una asignatura cuyo sistema de evaluación exige la redacción y presentación de una tesina final articulada en torno a los ejes temáticos del curso: el Derecho Constitucional de los Estados Unidos, con atenciones comparadas a otros ordenamientos. Tomando como base mis estudios previos sobre los modelos estadounidenses y las condiciones reales en que puede desplegarse una normativa de esta índole en una facultad de Derecho, elaboré un primer borrador.

Sometido dicho borrador al dictamen de colegas académicos, el texto reveló la necesidad de sucesivas revisiones (el cauce ordinario para la maduración de estos documentos). En mi condición de titular de la asignatura, asumí la toma de decisiones para transitar desde marcos teóricos abstractos hacia una arquitectura verdaderamente operativa. Finalmente, sometí la última versión consensuada entre docentes a la revisión de un sistema computacional, analizada a la luz de los parámetros de mi estudio sobre las políticas de IA en las facultades de Derecho estadounidenses. Utilicé para ello Harvey AI, dado que la facultad a la que pertenezco suscribió un convenio con sus desarrolladores. El resultado fue de una nitidez implacable: los sistemas informáticos debidamente adiestrados no se andan con contemplaciones con el ego del autor. Consideración que recibí con agrado, pues sirvió para recordar que en la técnica de redacción normativa siempre se es un aprendiz; que la claridad, aun en textos en apariencia nítidos, resulta esquiva; y que a menudo son las zonas de penumbra deliberada las que abren las vías más eficaces para una aplicación equitativa de la norma.

Con estos presupuestos, redacté un modelo de directiva sobre el uso de herramientas de IA para la asignatura. Dicha plantilla se nutrió de seis principios rectores que venía aplicando en años anteriores como régimen por defecto (prohibición absoluta de uso).

El modelo responde a necesidades formativas muy concretas: autorizar el empleo de AI Tools únicamente para dos fines específicos. El primero, la fase preparatoria y de acopio de fuentes previa a la redacción; el segundo, ofrecer un marco de auxilio a aquellos alumnos que piensan y redactan con mayor fluidez en una lengua distinta del inglés, permitiéndoles traducir posteriormente su trabajo para la entrega formal. El alcance de estos permisos puede ampliarse o restringirse según el criterio de cada docente mediante la simple modificación del apartado correspondiente del texto normativo.

A continuación se expone la fundamentación doctrinal de la directiva, seguida del resumen ejecutivo de sus reglas y del texto regulatorio en su integridad. Expreso mi agradecimiento tanto a las personas como a los sistemas informáticos que contribuyeron a perfilar este documento, el cual se concibe como un proyecto en continua evolución que se irá ajustando en función de los datos de su aplicación práctica con el alumnado.

 

PARTE 2: Fundamentación Doctrinal y Justificación de la Norma

1. Sobre la oportunidad de autorizar el uso de la IA en lugar de disponer su prohibición absoluta

Las herramientas de IA generativa se encuentran integradas en las plataformas de investigación jurídica, en los procesadores de texto y en el ejercicio profesional al que se incorporará el alumnado. Una prohibición categórica resultaría difícil de fiscalizar, desatendería la realidad operativa de la abogacía y desaprovecharía la oportunidad de instruir al estudiante en el uso responsable de estas tecnologías bajo la tutela académica. El régimen general del artículo 1 mantiene la exigencia de que el trabajo sea de autoría propia; el artículo 2 constituye una excepción delimitada y condicionada a dicha regla, no una licencia general de uso. Asimismo, su estructura modular permite al docente adaptar sus parámetros a los objetivos pedagógicos de cada curso.

2. Sobre la restricción de la excepción al trabajo de investigación final

Restringir el uso permitido de la IA a la tesina o trabajo final —prohibiéndolo de forma absoluta en exámenes, casos prácticos y ejercicios de aula— preserva un conjunto de evaluaciones continuas como parámetro objetivo para comprobar si el estudiante es capaz de desarrollar el análisis jurídico por sí mismo. El trabajo final recibe un tratamiento diferenciado por elaborarse a lo largo de un período más dilatado, lo que ofrece márgenes para la verificación transparente mediante apéndices y la tutoría continuada del docente, garantías inviables en una prueba presencial con límite de tiempo. Ello no obsta para que cada profesor pueda ampliar el ámbito de esta excepción dentro de los márgenes previstos por la normativa universitaria.

3. Sobre la preferencia de un catálogo ilustrativo frente a las reglas de trazo grueso (bright-line rules)

Trazar una frontera rígida y mecanicista entre la «asistencia autorizada» y la «sustitución ilegítima» exigiría disgregar la investigación, el análisis y la redacción en fases independientes e insostenibles en la práctica. Semejante pretensión en el trabajo cognoscitivo entraña dos riesgos: (1) generar la falsa confianza en el alumno de que todo lo no prohibido expresamente es legítimo, y (2) producir un efecto paralizador (chilling effect) sobre usos docentes plenamente válidos cercanos a la frontera reglamentaria. Por ello, los apartados i) y ii) del artículo 2.b) se configuran mediante listas ejemplificativas (a título enunciativo, no limitativo), manteniendo cierta flexibilidad interpretativa en los márgenes. Esta penumbra deliberada desanima la búsqueda de resquicios por parte del alumno y atribuye la facultad interpretativa al docente en el marco de las exigencias de su asignatura.

4. Sobre la exigencia de límites cuantitativos estrictos para citas y paráfrasis

El artículo 2.b.iii) reconoce que el material generado por IA puede incorporarse al trabajo como fuente secundaria: mediante cita textual o paráfrasis debidamente acreditada. Ello encauza el contenido sustantivo de la IA bajo las mismas exigencias de atribución aplicables a las fuentes humanas. A diferencia de los catálogos ilustrativos de los apartados i) y ii), esta cláusula impone topes numéricos fijos (máximo de 100 palabras por cita o paráfrasis y un tope del 10 % del cómputo total del trabajo por herramienta) en lugar de emplear fórmulas indeterminadas como «porciones reducidas». Aunque un criterio cualitativo como «reducido» puede funcionar entre profesionales, para el estudiante resulta ambiguo. Un límite numérico exacto aporta certeza previa al alumno y facilita la verificación posterior por parte del profesor.

5. Sobre la obligatoriedad de la transparencia, el apéndice documental y la atribución

Exigir la identificación de cada herramienta, la aportación del historial secuencial de instrucciones (prompts) y respuestas, y la acreditación formal en el texto responde a tres fines:

1. Crea un registro contemporáneo que permite al profesor auditar el uso efectivo del sistema frente a los estándares de la asignatura.

2. Consolida la transparencia como una pauta de conducta profesional, en consonancia con las exigencias deontológicas del ejercicio de la abogacía.

3. Invierte la carga práctica de la prueba: el estudiante que cumple rigurosamente reduce el riesgo de infracción, mientras que quien incurra en un uso no autorizado se enfrenta a un registro documental que dificulta el ocultamiento.

6. Sobre la función disuasoria de la declaración jurada de autoría

La firma de la declaración jurada convierte los deberes de transparencia en una manifestación formal de voluntad. Su falsedad constituye una falta independiente contra la honestidad académica, con sustantividad disciplinaria propia y al margen de la infracción de las reglas de uso.

7. Sobre la consideración del incumplimiento como uso no autorizado

El artículo 2.f) establece que el cumplimiento parcial no exime de responsabilidad. Un sistema basado en la transparencia solo es viable si las obligaciones de documentación se consideran constitutivas de la propia autorización. Permitir la elusión del apéndice o de la declaración sin perder la protección de la excepción desmoronaría la eficacia de la norma.

8. Sobre las potestades de inspección de la institución

El artículo 2.f) faculta a la cátedra y a la Facultad a revisar el apéndice documental de cualquier trabajo, exista o no sospecha previa de infracción. Un esquema basado en auditorías funciona como un elemento de disuasión continua, incentivando la veracidad del registro durante todo el proceso de redacción.

9. Naturaleza del texto como propuesta abierta

La presente plantilla se concibe como una primera aproximación sujeta a verificación empírica. Su diseño está previsto para someterse a prueba en la docencia real y perfeccionarse en función de la experiencia, sin pretender agotar con la mera redacción la resolución de los problemas éticos planteados.

PARTE 3: Síntesis Operativa de las Reglas para el Alumnado y Guía Práctica de Aplicación

A continuación se ofrece una explicación en lenguaje claro sobre la aplicación práctica de la norma en la asignatura. Este resumen parte del supuesto de que la única excepción a la regla general rige para el trabajo de investigación final. No añade obligaciones adicionales a las dispuestas en los artículos 1 a 6 y tiene carácter puramente orientativo:

1. Regla general: Salvo indicación expresa de la cátedra, todo trabajo evaluable debe ser de autoría propia y exclusiva del estudiante. Este es el presupuesto de partida de la normativa.

2. Exámenes, casos prácticos y ejercicios de aula: Queda prohibido el uso de herramientas de IA para redactar o modificar estas pruebas, salvo por las herramientas ordinarias de edición e investigación del punto 5 o para la traducción prevista en el punto 4.

3. Trabajo final de la asignatura: El estudiante podrá emplear herramientas de IA en la preparación de la tesina o trabajo final únicamente si cumple la totalidad de las siguientes condiciones:

   • Dirección intelectual: El análisis, la argumentación y las conclusiones deben ser de autoría propia. La IA se limita a fases preparatorias (brainstorming, comprobación de citas, sugerencia de contraargumentos y revisión de estilo). Se prohíbe generar análisis sustantivos o texto adoptado con meros retoques cosméticos.

   • Régimen de citas: Todo texto de la IA incorporado directamente al trabajo se citará como fuente secundaria en nota al pie y bibliografía, sujeto a los límites de 100 palabras por cita y al 10 % de la extensión total del texto.

   • Identificación: Se consignará el nombre comercial, versión y modo de acceso de cada herramienta.

   • Apéndice documental: Se adjuntará un anexo con el registro íntegro y ordenado de prompts e instrucciones enviadas y respuestas obtenidas.

   • Atribución y declaración jurada: Se incluirá la mención de atribución en el texto y se firmará la declaración del artículo 6.b.i.

   • Efecto del incumplimiento: La inobservancia de cualquiera de estos requisitos invalidará la autorización, estimándose el uso como no permitido.

4. Traducción de trabajos: Se permite traducir al inglés borradores redactados en otra lengua mediante IA o traductor humano, debiendo declarar el procedimiento, conservar el texto original y firmar la declaración del artículo 6.b.ii. El uso de traductores humanos requiere autorización previa y por escrito de la cátedra.

5. Herramientas de edición y bases de datos jurídicas: Los correctores ortográficos o gramaticales estándar y las funciones de las bases de datos de jurisprudencia (resúmenes o fichas) son de libre uso y no exigen apéndice documental, siempre que el alumno no copie directamente dichos textos y los redacte de forma autónoma.

6. Deber de consulta previa: Ante la duda sobre la legitimidad de un uso determinado, el estudiante debe consultar con el profesor antes de entregar la actividad.

7. Régimen disciplinario: El uso no autorizado de IA podrá ser tramitado como falta contra la honestidad académica conforme al Código de Honor, pudiendo acarrear la reprobación, suspensión o expulsión, así como su comunicación a las autoridades del bar u órganos de colegiación profesional.

    

TEXTO 4: Normativa Oficial y Directiva de la Cátedra se detalla a continuación

foto crédito aquí

* * * * * *  

Here are the links to the four parts of this study:    

1. Structure, Opacity, and Convergence: A Consolidated Analysis of Law School Generative AI Coursework and Exam Policies (10 July 2026) Larry Catá Backer (白 轲); SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7105978

2. "AI assists. You think. You analyze. You write. You take responsibility": Creating a Course AI Use Policy Template --Policy Text, Justification and Rule Summary for My Law & Religion Class at Penn State Dickinson (20 July 2026) Larry Catá Backer (白 轲); SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7187159

3. Rethinking AI Governance in Legal Education -- Five Machines (Grok, Harvey, ChatGPT, Claude, and Gemini), One Question, No Consensus but Five Archetypes： The Guardian, the Balancer, the Honest One, the Engineer, and the Philosopher on What Law Schools Should Do About AI （14 July 2026）Larry Catá Backer (白 轲) SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7119539

4. Part 3: Structure, Legitimacy, and the Limits of Machine-Centered Derivation: An Analysis of Five AI Systems’ Third-Stage Attempts to Construct Machine-Centric Governance Policies for Legal Education (26 July 2026); Larry Catá Backer (白 轲) (collaborating with HarveyAI, Claude, Gemini, Grok, and ChatGPT) SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7187159